One of my favourite bits of DMing advice comes, in all places, in the 2nd edition DMG. Zeb Cook (I am paraphrasing because I don't have it t...
One of my favourite bits of DMing advice comes, in all places, in the 2nd edition DMG. Zeb Cook (I am paraphrasing because I don't have it to hand) basically foreshadows some OSR thinking in one place only, which comes in his discussion of stat requirements for character classes. Don't let players access any class they want, he recommends; keep the stat requirements for rangers, paladins and so forth strictly. Encourage players to play the stats they are given, so to speak. Let the dice fall where they lay; so your PC didn't end up getting the 17 CHA required to be a paladin. Stop being titty-lipped about it - you can play a fighter who always wanted to be paladin but failed (or hasn't yet succeeded).
This post, and it’s comments, gives an interesting picture of one of the transitions in D&D: This advice applies far better to early D&D than to the direction publishing took in the second edition era. Melan, in the comments, puts it like this:
It could be done better. In the editor to Dragon Magazine #198 (An Extraordinarily Ordinary Hero), Roger Moore tells the story of Thora, a completely average fighting woman who got the deed to Tegel Manor and took it over with her merry band of friends, accomplices and hangers-on. But 2nd edition AD&D didn't do that. Which brings us to the second point. To make this work, you must not overshadow the characters. Yet 2nd edition was pretty much the Special Snowflake Edition: it was enamoured with characters with special destinies, elaborate class options (kits), and all the advanced classes. The rules that accumulated around the core system and the supplements (world guides and adventure modules) were invariably about this "other AD&D", the game of Big Special Heroes with powerful extra abilities and magic items. It told you your average joe character would be cool, then paired him up with Drizzt and Elminster and some guy from Planescape who considered your game world a tepid backwater, the Flyover Yokel Dimension (yes, that happened, and I am still kinda bitter about it). It told you a 18/00 Strength didn't matter, then gave you NPCs and sample characters with monstrously high stats until you said screw it and begged the DM for your gauntlets of ogre power. It told you not to get decked out as a magical Christmas tree, then put you up against other characters who were overloaded with expensive magical bling. It told you the little stories mattered, then laughed and talked about how cool those big, world-shattering stories with elaborate BBEGs were.
This is, in my estimation, true of the state of AD&D publishing during the 2e era. Official characters had attributes well above the curve for random generation. I can demonstrate this statistically using as my dataset the list of companions in the Baldur’s Gate series, which I’ll treat as representative of 2e-era published NPCs, but I’ll stick it under a cut because I imagine most of you don’t want to read several paragraphs of math.
3e chose to take this in a direction of casting aside the last vestiges of the idea that player characters are fundamentally ordinary people in extraordinary circumstances (and other editions of Wizards D&D continued this), but the roots of that started in the 2e era.
Arguably they go back further. The combination of defining attributes permanently at character creation (Tomes to permanently increase a stat by 1 being extremely rare, as were Wishes which could do the same), combined with having special rules which could only be invoked by characters with exceptional stats (classes beyond the basics, dual-classing) goes all the way back.
One could also argue that a system that purports to be about ordinary people, then kills your characters until you roll someone extraordinary, does not truly handle ordinary characters.
This is part of why I think the Wizards D&D idea of including attribute progression is a good one: it gives characters who start off ordinary the opportunity to nonetheless use rules that require high attributes, to gain legendary strength, and so on, though perhaps not to the extent that 5th edition does it where ability scores have stopped measuring potential and instead add directly to die rolls to the point that every monster needs to have a full attribute line.
